Bruno Abreu Calfa Email & Phone Number

East Hartford, Ct, Us

Arlington, Va, Us

Cultivated internal and external customer relationships as a Principal Investigator (PI) to start new projects and continue existing ones. Developed and implemented:- Supervised and semi-supervised (deep) learning algorithms (PyTorch, TensorFlow, Hugging Face) coupled with computer vision (OpenCV) techniques for robotic automated inspection.- Machine learning frameworks for sensor fusion of smart operations of airplane cabin areas (XGBoost, scikit-learn) and feature selection for conditioned-based maintenance (XGboost with NVIDIA GPU support and cuML).- Model calibration frameworks using field data via Bayesian updating methods (Approximate Bayesian Computation using PyMC for problems with expensive and black-box likelihood).- Simulation and optimization computational approaches (centrality-based using NetworkX) for supply chain management.

Arlington, Va, Us

East Hartford, Ct, Us

Project Management===============Served as a Principal Investigator (PI), Individual Contributor, and led team of Senior Research Engineers and UX/UI designers to develop and deploy customer-focused web-based tools (front-end: JavaScript, jQuery, Ajax, D3.js; back-end: Python (flask) and R)Experience with agile project management methods (e.g., Kanban board) as well as tools (e.g., Jira). Machine Learning=============Conceptualized, supervised, and implemented data analytics workflows involving the following steps and tools:(1) ETL process (PostgreSQL, Hadoop/Hive, and Python)(2) data profiling and cleaning (pandas_profiling)(3) feature engineering (by proposing new features and using insights from subject-matter experts)(4) feature selection (wrapper, e.g. XGBoost gain, and filter, e.g. joint mutual information, methods)(5) data visualization (interactive Bokeh plots, HoloViews, and Jupyter Notebook)(6) manifold learning via nonlinear dimensionality reduction (e.g., MDS)(7) dataset balancing (e.g., SVM-based SMOTE)(8) supervised learning (classification and regression; explored interpretability vs. accuracy) with hyperparameter optimization.Statistics=======Performed analyses and developed solution strategies for time-to-event or survival models (R and Python). Applied both frequentist and Bayesian methods (including multilevel or hierarchical models).Software Development=================- Improved different aspects of simulators developed in-house (C/C++). Designed and implemented simulation experiments to validate theoretical results.- Developed a model-in-the-loop (MIL) simulator, featuring state machines, using a model-based design (MBD) approach and prepared software requirements document (IBM Rational Rhapsody)- Web-based tools using Python (flask) and Ajax technologies

Madison, Wi, Us

My postdoctoral work is two-fold: chemical engineering education and Process Systems Engineering (PSE) research. In the education front, I am working with Dr. William F. Banholzer on enhancing the chemical engineering curriculum by developing business simulations designed to reinforce how the fundamentals of mass/energy balances, process flow simulation, and other engineering practices define successful commercial chemical processes. Regarding technical research, I am working with Dr. Christos T. Maravelias on PSE-related projects.

Pittsburgh, Pa, Us

My Ph.D. research can be broadly classified as Enterprise-Wide Optimization (EWO), which is a subarea of PSE. Specifically, I have worked on production scheduling and planning of chemical plants with batch and continuous processes. A significant part of my work has been on the development of data-driven methods for optimization under uncertainty. These methods rely on probability and statistics concepts to assist in uncertainty modeling and quantification frameworks.

Midland, Michigan, Us

Data mining and modeling to integrate historical and forecast data in the development of plausible planning scenarios.

Midland, Michigan, Us

Continued my Ph.D. research, collaborated with the Process Optimization group under Engineering and Processes Sciences (E&PS) R&D and the Advanced Analytics team.

Doctor Of Philosophy (Ph.D.), Chemical Engineering

B.Sc., Chemical Engineering